Overlapping community structures can reflect the fact that an individual can belong to more than one community, especially the individuals in social networks. It is important to detect overlapping community structures in online social networks. In this paper, an overlapping community detection algorithm is proposed for online social networks. Different from previous works, this algorithm can be applied in both undirected and directed networks. Moreover, for people who have accounts in several online social networks, this algorithm can be used to combine community structures of different networks. We assume that users' community structure in an online social network is actually a reflection of users' community structure in real life. By combining different reflections with the proposed algorithm, detected community structure can match users' interest-based communities better.
展开▼
